Diagnosing Specific Faults Through FRA Signature Analysis
The FRA Signature as a Diagnostic Fingerprint
The frequency response signature obtained from a Transformer Frequency Response Analyzer is a highly detailed representation of the transformer's internal electrical network, which is directly influenced by its physical geometry. Unlike routine insulation tests, FRA provides direct insight into mechanical condition. A stable, repeatable signature indicates mechanical integrity, while deviations from a trusted baseline fingerprint signal specific types of faults. Expert analysis focuses on changes in the pattern's resonant frequencies, the magnitude (in dB) of its peaks and valleys, and the phase response across the spectrum. The diagnostic process systematically correlates these changes with known failure modes.
Identifying Axial Winding Displacement
Axial displacement, often caused by electromagnetic forces during short-circuit events, results in a change in the winding's self-inductance. In the FRA magnitude plot, this fault primarily manifests as a horizontal shift of resonant frequencies at the lower and middle frequency ranges (typically between 1 kHz and 500 kHz) in the end-to-end open circuit test configuration. The entire curve pattern appears similar but is shifted along the frequency axis. A shift to lower frequencies suggests the winding has been physically elongated, increasing inductance. A shift to higher frequencies indicates axial compression. The phase plot will show corresponding shifts, providing confirmation of the inductive change.
Detecting Radial Deformation and Buckling
Radial deformation, such as bending or buckling of winding discs, alters the inter-winding and inter-disc capacitances. This fault is most clearly identified in the capacitive inter-winding test configuration and affects the middle to high-frequency range (from around 10 kHz to 1 MHz). The key indicator is a significant change in the amplitude (dB level) of resonant peaks rather than a simple frequency shift. The curve shape becomes distorted, with existing peaks dampened or new resonant peaks appearing. This reflects a change in the capacitive network within the winding. Severe radial deformation can also cause observable changes in the end-to-end short circuit test due to altered leakage inductance.
Recognizing Core and Clamping Structure Issues
Problems with the core and the mechanical clamping structure impact the magnetic circuit. These issues are prominently visible in the very low-frequency region (below 2 kHz) of the end-to-end open circuit test. A loss of core grounding or a change in the core's magnetic path can cause a large deviation in the magnitude response at these low frequencies. Similarly, a reduction in the clamping pressure on the windings and core can lead to a general increase in the magnitude response across a wide frequency band, as the overall mechanical damping of the system is reduced. Distinguishing between core and clamping issues often requires correlating FRA data with other tests like core excitation current measurement.
Systematic Approach to Comparative Analysis
Accurate diagnosis requires a strict, systematic comparison protocol. First, the current FRA traces must be compared against the transformer's own baseline, preferably the commissioning measurement. Second, a phase comparison (often more sensitive than magnitude) should be conducted. Third, results from all test configurations (open circuit, short circuit, capacitive, inductive) must be reviewed together to triangulate the fault location. Finally, the deviation should be quantified using statistical indicators like Correlation Coefficient (CC) or Root Mean Square Error (RMSE), as suggested by IEEE and IEC standards, to move from subjective visual assessment to objective, repeatable criteria. This multi-step analysis is crucial for distinguishing significant faults from acceptable aging or minor measurement inconsistencies.
Mastering the interpretation of FRA signatures transforms raw data from a Frequency Response Analyzer into actionable intelligence. By understanding the specific correlation between curve deviations and physical faults, maintenance engineers can precisely identify issues like axial displacement, radial deformation, and core problems, enabling targeted interventions that prevent minor damage from escalating into catastrophic transformer failure.
